To our knowledge the only other study in which

Hens that experienced a downshift in food reward took longer to reach the food than before the shift, but their postshift latency did not differ from that of control AHU-377 that had experienced the less preferred food throughout. The change in latency occurred over a few trials, and was thus in line with the Thorndikian law of effect (1911), which predicts a more gradual change in behaviour if there has been a strong reinforcement history, and matches similar response patterns shown in other experiments which have also shown a gradual decline in behavioural responses (Papini, 1997 and Petherick et?al., 1990). Following reward alteration, there was also a significant decrease in the food consumption of contrast birds, which reached but did not undershoot control group values. In contrast to the change in latency, this was an immediate effect seen in the trial in which the downshift occurred. An immediate decrease in food consumption was also seen in starlings, but it undershot control values indicating SNC (Freidin et al., 2009). Our food consumption result suggests that the chickens learned about, and were responding to the change in, reward properties (a prerequisite for SNC effects) rather than the behaviour simply being strongly associated, but there was no other evidence of an SNC effect.